MANTIS: An online, groundwater nonpoint source pollution decision support system

Abstract:

Nonpoint source (NPS) contamination of groundwater is a widespread and alarming environmental issue. Due to the (multi-) decadal time lag between management actions and groundwater system response (baseflow, well water quality), decision-makers rely on simulation tools that assess future water quality improvements. Existing contaminant simulation approaches are often ill-suited for simulating (diffuse) nonpoint source pollution at sufficient resolution of source and aquifer variability across large regions, and are too computationally expensive as decision-support tools. We present an alternative approach for an online decision support tool. It runs a rapid assessment of long-term future contaminant loading in thousands of wells based on management decisions made now. Users of the application make changes to cropping patterns and contaminant intensity. The model uses the inputs to adjust NPS pollution in a user-selected region and projects annual contaminant loadings in wells and streams for the next 100 years, factoring in historical conditions. The tool produces results in seconds, visualized as percentile curves of contaminant loading in wells and streams, allowing the user to react and explore many management scenarios and potential outcomes. The core of the tool is a detailed physically based groundwater flow and contaminant transport simulation model, using a mix of precomputation and validated simplifying assumptions to estimate resulting breakthrough curves quickly. In effect, the 3D modeling domain is decomposed into multiple 1D contaminant transport subproblems. Each transient subproblem (streamline) is solved for a unit input mass over a unit period as a source loading function to obtain unit response functions. The model computes hundreds of years of concentration breakthrough at receiving wells and streams by convoluting the unit response function with the user defined future NPS loading scenarios across a region and historical NPS loading. The tool has been applied and validated for the Central Valley, California, with over 20,000 large production wells and nearly 150,000 domestic wells. Results indicate that the tool is able to efficiently evaluate historical and alternative land management scenarios and their effect on nitrate in well water.